Introduction

Metal matrix composites (MMC’s) like aluminum matrix composites (Al MMC’s) and titanium metal matrix composites (Ti MMC’s) offer several unique advantages that their metallic counter parts lack. Wear resistance is chief among them. In fact, the presence of particles such as silicon carbide (SiCp) offers very high wear resistance for light metals such as aluminum and titanium. On the contrary, the presence of these second-phase particles is known to be detrimental to corrosion resistance (Ref 1, 2). Therefore, for those applications requiring both wear and corrosion resistance, the presence of these particles is not an ideal condition. For some applications such as brake rotor disks, the distribution of second-phase hard particles needs to be distributed in a graded manner such that components made out of the composites have differential properties. For example, a brake rotor disk of an automobile needs to posses high wear resistance at the outer periphery and good toughness at the inner core. Such a requirement can be readily met by functionally graded materials (FGMs) (Ref 3). In FGMs, composition and/or microstructure is varied along desired direction for which graded property (properties) is (are) required. Gradation in composition and/or microstructure can be continuous, as in the case of centrifugal casting (Ref 3, 4) or in discrete steps as in the case of graded composites produced by powder metallurgy route where in layers of different chemistry are built (Ref 5). It also needs to be pointed out that even chemical vapor deposition is proposed as a technique for production of some graded components (Ref 4). However, among these production techniques, centrifugal casting method is least expensive and is widely used for manufacturing graded rotor brake disks even though the powder metallurgy route is expected to provide disk having better mechanical and corrosion properties as disks produced by powder metallurgy route are superior in chemical homogeneity than those produced by centrifugal casting. Rajan et al. (Ref 3) have made the Al-SiCp functionally graded metal matrix composites (FGMMCs) by centrifugal casting method with 23 µm SiCp and examined the effect of centrifugal casting on the microstructure and on the mechanical properties, namely hardness and tensile strength. It has been reported by them that the concentration of SiCp was higher at the outer periphery of the casting and it decreased toward the inner core of the casting due to the centrifugal force. Further, as desired, the heat-treated disk showed gradual decrease in the hardness from periphery to the core of the casting. El-Galy et al. (Ref 4) have studied the effect of (a) size (16, 23 and 500 µm) and (b) concentration (2.5, 5, 10 and 15 wt.%) of SiCp on the microstructure and mechanical properties of Al-SiCp FGMMCs. According to them, smaller the particles higher is their ability to reach out to the outer surface of the disk as higher centrifugal force is required for large particles to transfer. Irrespective of the particle size, the outer periphery region showed higher tensile strength and hardness than the core of the casting.

Corrosion resistance is yet another important property of FGMMCs required for reliable performance of components, especially brake rotor disks. Brake rotor disks often come in contact with the road deicing salts, which are corrosive in nature causing premature failure (Ref 6). Only a limited number of studies are reported in the literature on the corrosion behavior of the so-called FGMMCs (Ref 5, 7, 8) although a few studies on the tribo-corrosion behavior of FGMMCs (Ref 7, 9) and corrosion behavior of Al-based normal MMCs have been reported (Ref 10,11,12,13,14). Velhino et al. (Ref 7) have studied the tribo-corrosion behavior of the Al/SiCp FGMMC using water (a mildly corrosive medium) as a lubricant in their tribo-corrosion experiments. This study showed that the wear rate of the composite increased due to water, as the presence of water helped in dislodging SiCp from the matrix during the wear process. Vieira et al. (Ref 9) have studied the influence of SiCp on general corrosion and tribo-corrosion in a functionally graded Al-SiCp composite. They employed unconventional Al-10 Si-4.5 Cu-2 Mg (wt.%) alloy to produce the FGMMC. Their study was related to the most exterior zone of the disk, where the area fraction of SiCp varied in the range 14-23% by area fraction. The authors reported that Ecorr is similar for all the samples, making them to suggest that the corrosion rate of the composite remains unchanged irrespective of the area fraction of the SiCp and heat treatment. However, no information on the icorr value, which is a direct measure of corrosion rate on any alloy, is available to confirm this suggestion. However, the repassivation tendency, which can influence tribo-corrosion, has been reported to depend on SiCp content. The repassivation tendency, revealed in a tribo-corrosion study, has been found to increase with SiCp content, indicating the beneficial effect of the reinforcement on tribo-corrosion. It is known that Ecorr of most aluminum alloys depends largely on cathodic kinetics, as these alloys are strongly passivating. The cathodic kinetics are much dependent on oxygen solubility in corrosive environment and can also vary if the corroding electrolyte over a metallic surface happens to be a thin layer as the migration of oxygen to the metallic surface becomes easier in such cases. Therefore, it is essential to understand the anodic polarization behavior of Al-SiCp composites to predict the corrosion tendency of composites under various prevailing conditions. While the reported literature on the corrosion behavior of Al-SiCp graded composite is scanty, such an understanding of anodic polarization, including the passivity of the composites has not been found in the published literature to the authors’ knowledge. Therefore, the present study concerns with understanding the corrosion behavior of a functionally graded A356-SiCp composite in 3.5 wt.% NaCl and to improve its corrosion resistance through heat treatment. Since the published literature on the corrosion of this system is scanty, the relevant literature on the corrosion resistance of normal Al-SiCp composite is presented here.

In general, composites are more prone to corrosion than the unreinforced alloys, due to chemical heterogeneities introduced by the reinforcements. A larger number of pits were reported to have formed in composites than in the corresponding unreinforced alloy (Ref 1, 13, 14). The pit density (number/area) was found to increase with increase in the volume fraction of reinforcement in the composites (Ref 14). This is because of the fact that the presence of reinforcement such as SiCp dispersed in the aluminum matrix disrupts the formation of a protective oxide film. Published literature showed that in SiCp-reinforced Al-Si-Mg alloys, interfaces of SiCp/matrix, Si/matrix, and intermetallics/matrix were the preferred sites for the localized attack. In A356 (Al-Si-Mg) alloy, these intermetallics, namely (β-Al5SiFe) phase, with platelet-type morphology, π-Al8Si6Mg3Fe phase, having Chinese script morphology were cathodic to the Al-matrix (Ref 15,16,17). Further, the eutectic Si precipitated out as a second phase during solidification producing a heterogeneous microstructure was also found to be cathodic to the matrix causing local galvanic attack (Ref 17,18,19,20,21). Arrabal et al. (Ref 18) using Kelvin probe technique reported that the interfaces between iron-containing intermetallics and eutectic Al phase act as initiation sites for pitting. Tahamtan et al. (Ref 21) reported that a reduced area ratio between the Si particles and the eutectic Al around Si particles could improve the corrosion resistance.

Pitting corrosion was reported to be a major form of corrosion affecting Al MMC’s (Ref 10, 11, 13). Trazaskoma (Ref 14) examined the effect of SiC whiskers (SiCw) reinforcement on the corrosion behavior of different Al alloy/SiCw composites, namely SiCw/A1 2024, SiCw/A1 6061 and SiCw/A1 5456 in the freely exposed and deaerated conditions in 0.1 and 0.6 N NaCl solutions. Pitting initiated even at corrosion potentials (Ecorr) in the freely exposed solution which was absent in the deaerated condition. A significant reduction in the oxygen content in the deaerated condition of the electrolyte drove Ecorr by several hundred mV toward active direction making the alloy resistant to pitting.

The pit morphologies in the SiCw/Al 6061 composite were influenced by the shape and distribution of SiCw (Ref 14). The pits were irregular in nature as the attack was mostly at the particle/matrix interface. On the contrary, the unreinforced alloy showed small and hemispherical pits, which can be attributed primarily to the absence of heterogeneous interfaces in the unreinforced alloy in comparison with the composite. Al/graphite MMC’s displayed higher corrosion rates than that of Al/SiCp MMC’s and Al/Al2O3 MMC’s (Ref 10, 22), possibly due to the fact that graphite is a good electronic conductor. The pits were deeper in SiCp-reinforced composites than in Al2O3 reinforced composites (Ref 23).

There is only a limited published work on the effect of volume fraction of reinforcement on the pitting as well as general corrosion behavior of Al-SiCp composites. Among the few studies found in the published literature, there seems to be no agreement on the role of reinforcement on the pitting potential (Epit) of composites. According to Buarzaiga et al. (Ref 19), an increase in SiCp content led to only a marginal increase in the Epit and a significant increase in overall corrosion damage was noticed. Contrary to this observation, Nunes et al. (Ref 23) and Pardo et al. (Ref 24) reported no significant change in the Epit and Ecorr values with increase in SiCp volume fraction, though the corrosion rate of the composites was found to increase with an increase in volume fraction of SiCp. According to Pardo et al. (Ref 24) study, Epit of the composite was affected mainly by the matrix composition and did not depend on volume fraction of the reinforcement of the composite. On the other hand, Feng et al. showed a decrease in Epit with an increase in SiCp volume fraction (Ref 25). Hence, it will be very useful to study how the volume fraction of SiCp reinforcement affects pitting as well as uniform corrosion behavior of the alloy.

The preceding literature review shows that only a few studies have been reported on the corrosion behavior of the so-called functionally graded Al/SiCp MMC Furthermore, no agreement exists on the effect of volume fraction of SiCp reinforcement on the uniform and localized corrosion. Therefore, the present study is undertaken to (1) understand microstructural variation including the SiCp volume fraction on the corrosion behavior of a centrifugally cast A356-SiCp composite disk and (2) improve its corrosion resistance.

Experimental Details

Materials

A FG A356-SiCp composite disk of 120 mm radius, 15 mm thickness as shown in Fig. 1 was centrifugally cast as per the details given in an earlier publication (Ref 3). Studies were carried out only on one disk. The SiC particles used as reinforcement have angular geometry with an average size of 23 µm. The measured density of casting was found to be 2.483 g/cm3, and the nominal chemical composition of the casting in wt.% is Si 7.0, Mg 0.3, Fe 0.2 max, Ti 0.2, Mn 0.1, Al (bal.).

Fig. 1
figure 1

Photograph of the FG Al-SiCp brake rotor disk studied. The depth over which the disk is expected to undergo tribo-corrosion (55 mm) and the deepest distance (65 mm) from the periphery of the disk up to which corrosion studies were made are marked

Full size image

Heat Treatment

Specimens at distance of 5, 25, 45, 65 mm starting from the periphery of the centrifugally cast disk were cut. Specimens obtained at each location were studied in (a) the as-cast condition, (b) solution heat treated at 540 °C for 2 h (just below the eutectic temperature of the base alloy to get maximum solubility of alloying element) and water-quenched to room temperature, (c) solution heat treated as (b) and subsequently aged at 175 °C for 2 h and then furnace-cooled to room temperature.

Specimen Preparation and Characterization

For microstructural characterization, the specimens were ground with silicon carbide abrasive papers of successive grades starting from 400 to 3000 grit. The specimens were electropolished at 18 V potential using 85% methanol and 15% HNO3 for 14 s. Optical (Olympus with Olysia m3 software) and scanning electron microscope (JEOL-7600F) were employed to examine the microstructures. The area fraction of SiCp was obtained by analyzing optical images at 100 × using OLYSIA m3 image analysis software.

Phase identification and compositional analysis of the samples were carried out using energy-dispersive spectroscopy (EDS) in SEM and x-ray diffraction (Cu Kα 1.540598 Å). A step angle of 0.01° and a scan speed of 0.15 s/step were used in x-ray diffraction studies.

Corrosion Tests

Electrochemical corrosion experiments were performed using a Gamry-600 computer-controlled potentiostat connected to a three-electrode flat bottom cell. Samples with an area of ~ 0.5 cm2 formed the working electrode, a Pt sheet was the counter electrode, and a saturated calomel electrode (SCE) was used as reference electrode. Studies were also carried out just below the wearable region, as crevice formed over there might affect the corrosion tendency of the disk. Working electrodes were prepared by mounting the samples cut at different distances of the disk and heat treated as described in the earlier section in cold setting resin with one end connected to a copper wire. The samples were then grounded with successive grades of SiC papers from 600 to 1200 grit size followed by ultrasonic cleaning in methanol. All the corrosion experiments were carried out at room temperature. The test solution was 3.5 wt.% NaCl. Electrochemical corrosion tests were carried out in deaerated condition by purging nitrogen gas for 1 h before the commencement of the experiments. Potentiodynamic polarization experiments were conducted in the potential range of − 1.1 to − 0.6 V (SCE) and at a scan rate of 0.5 mV/s. Potentiostatic experiments were performed at a fixed potential of − 0.9 V versus SCE for 4 h. All the electrochemical corrosion tests were repeated at least 3 times.

Immersion tests were carried out on samples at 5 and 65 mm distances, respectively, from the outer periphery of the disk. The samples were immersed in 3.5 wt.% NaCl solution freely exposed to the atmosphere for 30 min. Post-immersion, the samples were ultrasonically cleaned in methanol. Nature of corrosive attack was observed using a scanning electron microscope.

Results and Discussion

X-ray Diffraction and Microscopy

Phase analysis and microstructures of the FG disk in as-cast and heat-treated conditions were examined at 5, 25, 45 and 65 mm distance from its periphery to illustrate possible variation in the nature and distribution of phases along the radius of the disk. X-ray diffraction patterns of the disk obtained at 5 and 65 mm distance from the periphery of the disk are shown in Fig. 2 and 3, respectively. The XRD patterns revealed the presence of SiCp, π-Al9FeMg3Si5 (hexagonal) and Mg2Si phases at the outer periphery, whereas close to the core of the disk (65 mm from the periphery) only π-Al9FeMg3Si5 and Mg2Si phases were found with no detectable SiCp. The β-AlFeSi could not be detected in the XRD pattern, but could be found in SEM, confirmed using EDS.

Fig. 2
figure 2

XRD patterns of graded A356-SiCp brake rotor disk at 5 mm distance for all the three conditions

Full size image
Fig. 3
figure 3

XRD patterns of graded A356-SiCp brake rotor disk at 65 mm distance for all the three conditions

Full size image

The distribution of these phases was analyzed using optical and scanning electron microscopes and energy-dispersive analysis. Typical optical microstructure of the as-cast A356/SiCp FG brake rotor disk obtained at 5, 25, 45 and 65 mm distance from the periphery is shown in Fig. 4. For heat-treated conditions, only the SEM microstructure corresponding to 65 mm distance from the periphery of the disk is shown here (Fig. 5b), as it showed a perceptible difference in the microstructures due to heat treatment. Various phases seen in SEM were identified based on their chemical composition obtained using EDS and mapping the chemical composition of the microstructural feature to the phases identified using XRD data. It can be said that the specimens examined at different locations showed the presence of all these phases detected by XRD. However, the extent to which these phases were present was found to vary from the outer periphery to the inner core. To illustrate this, elemental mapping of specimens taken at 5 and 65 mm distance from periphery is shown in Fig. 6 and 7. Variation in SiCp area fraction from the outer periphery to the core of the disk is shown in Fig. 8. It is necessary to mention that the contrast for the SiCp in relation to the eutectic silicon has been poor when examined in optical microscope (Fig. 4). However, an electropolished specimen seen in SEM nicely revealed these particles as brought out in the backscattered image of SEM (Fig. 9a). As shown in Fig. 8, the SiCp area fraction was the highest at the outer periphery and gradually decreased toward the inner core. On the contrary, the amount of eutectic phase and intermetallics phases (β and π) was less at the outer periphery than at the inner core (Fig. 9). The size of intermetallics is smaller at the outer periphery and increased toward the center of the disk. However, the outer periphery has more number of intermetallics than the core of the disk. The reduction in size of the intermetallics at the periphery can be attributed to the hindrance offered by the SiC particles (Ref 19) which are more in number at the periphery than at the core.

Fig. 4
figure 4

Microstructures of the as-cast A356-SiCp brake rotor disk at: (a) 5 mm marker-1 shows SiCp; (b) 25 mm; (c) 45 mm and (d) 65 mm. Intermetallics, eutectic Si and SiCp, respectively, are shown with markers bearing numbers 1, 2 and 3

Full size image
Fig. 5
figure 5

Microstructure obtained at 65 mm distance showing phases in (a) as-cast condition. Arrows marked as 1, 2 and 3 are showing eutectic Si, β phase, π phase, respectively, (b) solution heat-treated condition where arrows marked as 1, 2 and 3 are showing π phase, β phase and modified Si, respectively (distances are measured from outer periphery)

Full size image
Fig. 6
figure 6

X-ray maps of (a) Al, (b) Si, (c) Mg and (d) Fe taken at 5 mm distance from the periphery of the FG A356-SiCp disk. Corresponding SEM image of the specimen is shown

Full size image
Fig. 7
figure 7

X-ray maps of (a) Al, (b) Si, (c) Mg and (d) Fe taken at 65 mm distance from the periphery of the FG A356-SiCp disk. Corresponding SEM image of the specimen is shown

Full size image
Fig. 8
figure 8

Volume fraction of SiCp vs. distance from outer periphery

Full size image
Fig. 9
figure 9

SEM micrographs of the as-cast A356-SiCp brake rotor disk at: (a) 5 mm (b) 25 mm; (c) 45 mm and (d) 65 mm highlighting the presence of various phases. Various phases SiCp, eutectic Si, β and π, respectively, are shown with markers bearing numbers 1, 2, 3 and 4

Full size image

Coming back to the role of heat treatment, as revealed by the SEM images of Fig. 5(a) and (b), respectively, for the as-cast and heat-treated disk, it can be said that the solutionizing treatment turned some of the eutectic Si needles of the as-cast disk into spheroids. It appears that the solutionizing duration is inadequate to transform all the Si needles into spheroids. As the needle-type Si is detrimental from mechanical properties point of view, the solutionizing heat treatment is expected to improve these properties. However, what is interesting is that the solutionizing heat treatment is expected to lower the segregation of elements, which in turn is expected to increase the corrosion resistance of the disk. The heat treatment has been found to affect only the eutectic Si as the intermetallic phases (β and π) are found unaffected. The published literature indicates that the iron-containing intermetallics (β and π) transform only above ~ 580-620 °C (Ref 26). Therefore, these phases are expected to be stable at the solutionizing temperature of 540 °C. Notably, the aging treatment is expected to result in the precipitation of Mg2Si phase in A356 alloy which was seen in the XRD pattern (though the peaks are less intense as the time for aging treatment was less, i.e., 2 h, which is in agreement with the earlier report (Ref 27). However, the presence of this phase could not be detected in SEM as they are expected to be smaller in size. Such precipitation is expected to improve tribological properties as it results in increasing the hardness of the alloy, although the same thing cannot be said about tribo-corrosion as the corrosion behavior of this phase is equally important to be considered in tribo-corrosion.

Corrosion Studies

Figures 10, 11 and 12 show the potentiodynamic polarization curves of the specimens of the disk obtained at 5, 25, 45 and 65 mm distances from the periphery in all the three conditions, namely the as-cast, solutionized and solutionized + aged. From these plots, the variation of the corrosion current density (icorr), passivation current density (ipass) and pitting potential (Epit) with respect to the distance in all the above is presented in Fig. 13, 14 and 15, respectively. The following points regarding the relation among corrosion, heat treatment and the position of the disk emerge:

Fig. 10
figure 10

Potentiodynamic polarization curves of as-cast FG A356/SiCp brake disk samples at different distances from outer periphery

Full size image
Fig. 11
figure 11

Potentiodynamic polarization curves of solutionized FG A356/SiCp brake disk samples at different distances from outer periphery

Full size image
Fig. 12
figure 12

Potentiodynamic polarization curves of solutionized + aged FG A356/SiCp brake disk samples at different distances from the outer periphery

Full size image
Fig. 13
figure 13

Variation of the corrosion current density of FG disk with distance for all the three conditions, namely as-cast, solutionized and solutionized + aged

Full size image
Fig. 14
figure 14

Variation of the passivation current density of FG disk with distance for all the three conditions, namely as-cast, solutionized and solutionized + aged

Full size image
Fig. 15
figure 15

Variation of pitting potential of the graded disk with the distance from the periphery to the core in all treated conditions obtained deaerated 3.5 wt.% NaCl solution

Full size image
  1. 1.

    The as-cast samples had the highest icorr and ipass values, the solutionized the least and the solutionized + aged samples lie in between for all the locations studied.

  2. 2.

    Excepting the solutionized condition, where no change is noticed, the icorr and ipass values of the disk increased from the periphery to the core in the order: 5 < 25 < 65 < 45 mm.

  3. 3.

    Within the observed scatter in the data, it can be said that the pitting potential increased in the order as-cast < solutionized + aged < solutionized.

The above-mentioned behavior of the alloy with respect to icorr and ipass can be explained as follows: Examination of the polarization curves shows that excepting for the specimen at 25 mm in the solutionized condition, the cathodic kinetics increases from the periphery to the core in all the cases irrespective of the heat treatment. The same thing cannot be said about the anodic kinetics. It appears that dissolution of the disk is the highest at 45 mm in the case of the as-cast condition as revealed by the ipass and to some extent the icorr. The fact that cathodic kinetics increases from the periphery to the core and the area fraction of the SiCp decreases from the periphery to the core (Fig. 10, 11, 12) of the disk indicates that SiCp is not an effective cathodic phase. So, it is not expected to affect the localized corrosion through galvanic interactions. As its presence also lowers the passive current density, it indicates the fact that it is inert in nature at least in the potential region over which the samples were scanned. As a consequence, the disk exhibits a broadly inverse relation with respect to SiCp content from the point of view of icorr as well as cathodic kinetics, especially in the case of as-cast disk. Considering the fact that the SiCp is electrochemically inert, they can play a significant role in lowering the gross electrochemical kinetics, which includes icorr and ipass of the composite. However, the fact that the disk at 45 mm showed notably higher icorr and ipass than the corresponding parameters at 65 mm distance, despite the former having higher SiCp than the latter, indicates that other factors also seem to influence the corrosion behavior, specifically the anodic kinetics of the disk. This aspect is examined later through the potentiostatic polarization studies. However, it should be pointed out that contrary to our observations Nunes et al. (Ref 23) and Pardo et al. (Ref 24) reported an increase in corrosion rate of composite with increase in SiCp content. Our work is also in disagreement with the Vieira et al. (Ref 9), who reported that corrosion rate of the Al-SiCp did not vary with SiCp content and/or heat treatment.

With regard to anodic dissolution, the disk seems to suffer the highest dissolution at 45 mm (Fig. 10, 11) in the as-cast and the solutionized conditions. However, aging has changed the trend. In this condition, the anodic dissolution of the disk at 65 mm distance has been found to be the highest. This was further investigated using potentiostatic studies. Potentiostatic curves of FG A356/SiCp disk held at − 0.9 V (SCE) in all the conditions, namely as-cast, solutionized and solutionized + aged are shown in Fig. 16, 17 and 18. The curves show that passive current density at the outer periphery was the least and the highest at the 45 mm distance, which complement the anodic polarization curves. The potentiostatic curves of Fig. 16, 17 and 18 show the anodic current densities of the specimens held in the passive region − 0.9 V (SCE). The fact that specimens at 45 mm show the highest anodic current even though its SiCp content is not the least indicates that the dissolution and passivation of the disk cannot be merely related to the area fraction of the SiCp. This becomes even clear due to the fact that the disk in the solutionized condition did not show any significant variation in ipass with the distance. It appears that chemical inhomogeneity plays a very significant role on the passivity and corrosion behavior of the disk. The nature of the corrosive attack of a specimen in the solution annealed condition post-potentiostatic tests is shown in Fig. 19. Microstructural examination post-potentiostatic tests confirmed that the attack started at Mg-rich regions and dissolution of the Mg-rich phases at the interdendritic regions occurred even in the passive range. The attacked regions are rich in Mg. Notably, intermetallics (β and π) are not attacked, which is better illustrated in Fig. 22. The extent of attack seen on the specimen has been found to be proportional to the anodic current observed in the potentiostatic tests, indicating the fact that the selective dissolution of these phases are responsible for the increase in the anodic attack. It also appears that these anodic regions/phases are unevenly distributed in the disk. Overall dissolution could be a summation of the currents from various microstructural constituents. This has been expected through examination of the post-corroded specimens. It is interesting to note that the anodic current density of the specimen taken at 5 mm distance continues to drift toward the lower value with time, indicating elimination of active area, by anodic dissolution, thereby making the surface more stable against the corrosion and tend to become passive.

Fig. 16
figure 16

Potentiostatic curve of as-cast FG A356/SiCp brake rotor disk in deaerated 3.5 wt.% NaCl solution at − 0.9 V vs. SCE

Full size image
Fig. 17
figure 17

Potentiostatic curve of solutionized FG A356/SiCp brake rotor disk in deaerated 3.5 wt.% NaCl solution at − 0.9 V vs. SCE

Full size image
Fig. 18
figure 18

Potentiostatic curve of solutionized + aged FG A356/SiCp brake rotor disk in deaerated 3.5 wt.% NaCl solution at − 0.9 V vs. SCE

Full size image
Fig. 19
figure 19

Nature of corrosive attack (a, b) and (c) x-ray mix map of solution annealed A356/SiCp disk in deaerated 3.5 wt.% NaCl solution at − 0.9 V vs. SCE post-potentiostatic test

Full size image

Specimens immersed freely in 3.5 wt.% NaCl solution exposed to air have been found to show a different behavior. Here, the attack was also seen at the interface of SiCp/Al and intermetallics/Al interface (Fig. 20). The reason for this variation can be attributed to the fact that in a freely exposed condition Ecorr of the specimen lies very close to that of Epit (anodically polarized). Similar studies have been reported by Trazaskoma et al. (Ref 11). Therefore, the passive film is expected break, leading to the attack at the interface. The passive film is expected to be less stable at the interface due to elemental segregation as well as the inability of the alloy to form a defect free film. Similar behavior has been reported by other authors as well (Ref 14, 24). Further, at 65 mm distance in both the as-cast (Fig. 21) and solution annealed (Fig. 22c and d) conditions, the Mg-rich region adjacent to the eutectic Si was attacked first and for comparison SEM image of the unexposed specimen corresponding to the identical location is shown in Fig. 22(a) and (b). The micrograph brings out the fact that the Mg-rich regions that lie adjacent to the eutectic Si and the intermetallic phases were selectively attacked. It is well known that Mg is relatively active in comparison with Al and Si, so Mg-rich area preferentially dissolved from the eutectic.

Fig. 20
figure 20

SEM image of solution annealed graded A356/SiCp brake rotor disk at 5 mm distance from outer periphery showing corrosive attack after 30 min of immersion in 3.5 wt.% NaCl solution

Full size image
Fig. 21
figure 21

SEM micrograph of as-cast graded A356/SiCp brake rotor disk at 65 mm distance from outer periphery showing corrosive attack post-immersion in 3.5 wt.% NaCl solution for 30 min

Full size image
Fig. 22
figure 22

SEM image (a) and x-ray mix map (b) of the solution annealed graded A356/SiCp brake rotor disk at 65 mm distance from outer periphery before immersion in 3.5 wt.% NaCl solution and SEM image (c) and x-ray mix map (d) of the solution annealed disk at 65 mm distance obtained after 30 min of immersion in 3.5 wt.% NaCl solution, arrows indicating the corrosion attack

Full size image

Notably, the Ecorr of the disk remained almost the same with respect to the distance. Since Ecorr depends on both the anodic and cathodic kinetics, simultaneous change in both the kinetics (as shown by the polarization curves) may result in the alloy having similar Ecorr values, although icorr is expected to change. From the location-dependent corrosion behavior point of view, the corrosion resistance of the disk increased in the order of 45 < 65 < 25 < 5 mm. Corrosion resistance of the disk at 25 mm distance is better than that of the disk corresponding to 65 mm distance from the outer periphery The reason why the disk at 65 mm shows the lowest icorr (highest corrosion resistance) than the disk at 45 mm distance may be attributed to the larger amount of intermetallics (β and π) and SiCp present at the former location, as all these phases are less prone to corrosion than the matrix. The above suggestion also arises out of the fact that the disk shows lower ipass at 65 mm than at 45 mm from the periphery. The effect of galvanic corrosion does not arise here as the alloy is forcefully polarized anodically through a potentiostat, an act far more severe than simple galvanic coupling aided by their respective galvanic/corrosion potentials. At 45 mm where the extent of both the SiCp and intermetallics was less, the alloy suffered the highest uniform corrosion as noted by the passive current density.

As the outer periphery of the brake rotor disk is expected to experience high wear during operation, the material over this area needs to possess high hardness. Interestingly, SiCp content does not affect the uniform corrosion resistance of the disk, although published literature indicates its detrimental effect (Ref 19). However, the Mg-rich regions have been found to be prone to selective dissolution. The effect of these Mg-rich regions is found to be more toward the core of the disk than at the outer periphery. The solutionizing and solutionizing + aging treatment improved the corrosion resistance more toward the core of the disk than at the outer periphery. The Epit values did not change with distance in all the conditions, namely as-cast, solutionized and solutionized + aged, which was in good agreement with the study by Pardo et al. (Ref 24). This probably happens due to the fact that more vulnerable regions (microstructures) of the alloy are expected to decide the Epit and any variation in the uniform corrosion is not expected to affect the Epit. Thus, solutionizing heat treatment, which is expected to lower chemical heterogeneity, has increased Epit. It also appears that since solutionizing can bring down Mg2Si phase, which is more prone to corrosion, it can also increase Epit. It is also necessary to point out the fact that while macrogalvanic cells are expected to operate on the disk due to the presence of heterogeneous phases and condensation of water, macrogalvanic corrosion due to the distribution of phases across the radius of the disk is unlikely to occur in the operating conditions of automobiles. This is because of the fact that disk is not expected to be wet continuously and moreover corrodent layer will be very thin limiting the current flow to the localized regions.

Conclusions

  • Corrosion resistance of centrifugally cast A356/SiCp was studied in 3.5 wt.% NaCl using potentiodynamic polarization technique and microscopy. Microstructures revealed that gradation in SiCp and intermetallics distribution along the radial direction.

  • Corrosion resistance of the composite was improved through (a) solution annealing and (b) solution annealing and aging. Between the two, the former has shown the highest corrosion resistance and passivity. The improvement in corrosion resistance has been attributed to the chemical homogenization of the as-cast structure due to solution annealing. A reduction in corrosion resistance and passivity due to the aging of the solutionized composite can be attributed to the formation of Mg2Si phase, which suffers severe corrosion. Since the wear resistance of this alloy depends mainly on the presence of SiCp, even in the solution annealed condition, the composite disk is expected to offer tribo-corrosion resistance at par than that of the solution annealed + aged condition (as aging increases the hardness of the composite, which is also a requirement for tribo-corrosion resistance).

  • The composite suffered preferential corrosion attack on the Mg-rich regions.

  • The corrosion resistance of the disk was found to vary in the order 45 < 65 < 25 < 5 mm in the as-cast condition. The passivation current density of the disk excepting one case (at 25 mm location solution annealed and aged) followed the same trend. Broadly, the presence of “inert SiCp” and relatively resistant intermetallic phases are responsible for these behaviors. However, selective dissolution of Mg-rich phase, whose quantity varies with distance, alters the trend set by inert SiCp and relatively resistance intermetallic phases.